No I don't think you did miss it. Peter mentioned above boards back in October. I just hope that things don't slip too far, because before you know it, Christmas break appears then CNY etc. The last couple months of each year seem to always speed up and rush by, it's like a time warp.
I've been in contact with publison to see if he is happy to handle postage of orders in the US not only of the P2D2 but also the other more general boards, some of which will be listed on Amazon (perhaps). P.S. He is btw
I'm finding out how much I can get the simple uPORT USB serial module assembled for and if the price is right I will sell them at the right price. Same goes for the P2D2 and the pricing will be aggressive, let's say, cost plus. The initial run will be done locally though.
There's the Moon festival in China at the moment adding a little delay to getting stuff back but in the meantime I have imported my files into Kicad and have even played with getting the 3D models working (mostly). Here is the P2D2 mostly right (the outside row of holes are castellated). (P2D2 QUICK REFERENCE)
Great news, Peter !
Please do not forget us, poor Europeans, and get in touch with @ErNa too . And an info on how to place the order wouldn't hurt as well.
PROTOTYPES
If all goes well my P2D2 boards and sundries will be here in a few days and I will be busy building up a tester. If I'm happy with the design and how the boards go together I will make up all I can with the stock I have but rather than ordering more parts at the moment I will get quotes on A&T from JLCPCB as they can also supply most of the parts from their LCSC division. Maybe I should give a US source a chance here too.
DESIGN
As you have seen, my designs are based on practical and commercial use, not educational nor hobbyist. Neither are they necessarily plug-in breadboard friendly although one can created a special breadboard JM style. My software is not purely for software purposes, but to run the hardware. My hardware is not ignorant of the software that runs on it either and even the layout reflects how I envisage the software will use it. P2D2 is a holistic design and while it may also look "pretty", it is entirely practical and versatile. At 2.4sq" including connectors it is a very compact design especially considering it is not a "bare-bones" design, but is fully fleshed out, and even more so with the P2PAL "layer" added. Placed next to a 40-pin DIP P1, there is not a lot of difference.
MOUNTING
When it comes to mounting the P2D2 can be mounted smd style flat to the pcb or 50mil matrix board with a 4mm max height, but even with 50mil pin header socket on the P2D2 component sided and flip mounted into pin headers that height is only around 6mm or 1/4". Of course you can also use 100mil pin headers when the P2D2 is smd mounted to the header card that has the same pinout and spacing as the original P2D2r4 along with the RPi header.
USB & RESET
Some may think that having USB onboard is great and some may think it is a hindrance and can cause problems with resetting etc. While that may be true of typical Parallax designs, I would never have USB if I couldn't bypass it and if it reset inadvertently either. In fact the USB is only part of the function of the support micro that is powered directly from the P2 supply since reset control is of paramount importance in a commercial design that needs to run 24/7 in a hostile environment. Besides the guaranteed reset that comes from a passive pull-down, active pull-up on power-good approach, there is also the watchdog that can at least restart the P2 and also maintain a non-volatile watchdog timeout log for fault analysis. BTW, the P2D2 will not reset simply because of USB or terminal connection/disconnect, it will only reset with a valid serial loader pulse, and even then it makes sure the P2D2 is forced into serial load mode despite the boot Flash and SD. The USB and reset support firmware is upgradable over USB and configurable in-circuit too. There is never any need for manual dip switches or jumpers. You can also drive P63/62 directly and besides the current limit resistor that makes that possible, the UB3 also floats its transmit line if it is not actively connected to USB.
SUPPORT FEATURES
Don't forget that while the P2D2 is plus featured, that it is still an economic design. The tight and optimized 2-layer board is very cheap to make, the USB and reset support only adds around $1 to the cost, and the same too with the flexible Si5351 clocking features where we can use precision crystals such 26MHz GPS crystals and still have a standard 20MHz, or any odd frequency up to 200MHz in fact. Then there are the multiple SD card footprints for various positions on the board along with software control over SD power cycling, but not only that, it also has series termination resistors for optimized high-speed drive. Real products normally need some kind of RTC, that is why may MCUs these days have a dedicated RTC built on-board, but nowhere near as good as an external RTC. The RV-3028-C7 is a 40na standby champion, and extremely accurate.
POWER
As well as switching the 1.8V supply, the 3.3V supply is also switched but at a slightly higher voltage and fed through a dual LDO for clean 3.3V I/O power. Besides the efficiency benefits of this approach it also benefits external circuitry since the 3.6V is available and can be fed directly into external chips without problem, even LDO regulated if needed. In fact the HyperRAM and ESP32 are fed directly from the 3.6V on the P2PAL layer which is especially important when you consider that the ESP32 may need 500ma during transmit. If a P2 system needed 1A on the 3.3V supply and it only used linear regulation then that would amount to 1.7W of heat dissipation no matter what, even if series diodes etc were used. A 3.7V Lithium can be connected to the VIO and there is a back-feed diode that ensures that the switcher is powered so that it can supply the 1.8V.
OPTIONABLE & UPGRADABLE
There are so many options to the base P2D2 itself but with extra P2PAL layer it is effectively turned into PC module with HyperRAM and the ESP32 etc, yet this thin layer added directly to the back of the P2D2 adds only 4mm to the height with 3mm of that being the ESP32 itself, and 1mm for the layer. But other P2PAL versions could be developed that mesh with the P2D2 that might have more memory or even a dedicated high resolution video CPU that can directly address video memory thus offload the problem of hires/hicolor displays and provide true USB host ports.
DELIVERY
This is the part that may be more interesting than even the compelling features of the P2D2 itself. If all goes well then October will see all the preorders filled and this week too I will take the time to test and update the working P2D2 boards I have that I had earmarked as free evaluation boards to some but due to the earlier problems with finalizing the UB3 firmware, I've had to belay that action while I make sure that everything else is sorted out (please give me a gentle pm nudge to remind me please). During the next few weeks as I have previously mentioned I will also find out about getting these made in an SMT assembly house, perhaps even JLCPCB/LCSC themselves, in which case there will be plenty of stock which will be kept locally both here and in the USA and also Europe I would imagine. btw, beyond these early production boards I have a very low and attractive target price set in mind that will make it a no-brainer to use P2D2.
SUPPORT
Yes please. I need your support.
Thank you
Just noticed those two LEDs on the less populated surface showing in the 3rd picture above might interfere with the SD card if their height is too high. It's not an issue for the one on the other side because the connector faces the other way towards the P2, so you may not have noticed this in the design. If you find you need a re-spin for any reason maybe rotate and move them down to under the USB socket, or spread one either side of the SD sockets if there is room perhaps... there's probably space to run the extra traces.
Just noticed those two LEDs on the less populated surface showing in the 3rd picture above might interfere with the SD card if their height is too high. It's not an issue for the one on the other side because the connector faces the other way towards the P2, so you may not have noticed this in the design. If you find you need a re-spin for any reason maybe rotate and move them down to under the USB socket, or spread one either side of the SD sockets if there is room perhaps... there's probably space to run the extra traces.
The flipside SD header is kind of redundant now as I find that if I do place a socket on the other side, I'd rather have it enclosed which is possible with the 3rd footprint which accommodates a fully enclosed push-push style socket. There are also these cheap naked but enclosed uSD sockets that Cluso99 sent me some of and they happen to fit on the flipside header pads but as a fully protected socket. So I don't think I'd ever use an actual Molex uSD header in that position whereas on the component side it is a real space-saver.
Just noticed those two LEDs on the less populated surface showing in the 3rd picture above might interfere with the SD card if their height is too high. It's not an issue for the one on the other side because the connector faces the other way towards the P2, so you may not have noticed this in the design. If you find you need a re-spin for any reason maybe rotate and move them down to under the USB socket, or spread one either side of the SD sockets if there is room perhaps... there's probably space to run the extra traces.
SD cards and LEDs have various height choices.
You can also get right angle LEDs, and those can avoid the SD card covering the main LED light of a vertical LED.
I like these ones, bright, and a good price, and not too small... - you can do a PCB design that takes a selection (vert or right angle) , just get the right angle ones right https://lcsc.com/search?q=LTST-S220
SD cards and LEDs have various height choices.
You can also get right angle LEDs, and those can avoid the SD card covering the main LED light of a vertical LED.
I like these ones, bright, and a good price, and not too small... - you can do a PCB design that takes a selection (vert or right angle) , just get the right angle ones right https://lcsc.com/search?q=LTST-S220
This is true too as I forget that these Molex headers are available in various heights for that reason. The right-angled LEDs can only be used when the footprint and the angle match the polarity. Mind you, the P2PAL layer could be used simply for the LED indicators and uSD pattern, minus the ESP32 and HyperRAM etc. That way there are 6 LEDs altogether plus the WS2812 as well.
The pcbs have been shipped and I'm expecting them this Friday or possibly Monday. I have switched over to Kicad now to finalize the P2LAB development board and other simpler pcbs. Here's one that I did for a DIP40 where the flipped P2D2 plugs into on top and then the machine pin socket or possibly pin headers on the other side. So it is a very cheap and simple adapter where one type is P1 footprint compatible and another type crams 38 I/O plus power&ground onto it. Same too with the RPI HAT.
btw, I will probably ship the blank DIP-40 and the QUIP-80 (original 100mil style P2D2) pcbs with each P2D2 as standard accessories along with the header pins. Mount them whichever way you wish then. I had thought of doing a P2PAL type layer that had the DIP40 socket for a very low profile connection but that would have precluded the use of an actual P2PAL enhancement. With a separate adapter it becomes an easy matter to swap around the P2D2 etc.
I didn't think I needed to show this but this is the 100mil header layer for the P2D2 module itself. The common P2D2 core module is surface mounted to the top of this layer, thus becoming one pcb, and there are pads underneath for a P2PAL to be mounted as well. The extra 40-pin header is for plugging directly into an RPi, although I feel that the P2D2-HAT is a better candidate for that.
btw, the original version was done in Protel but this is a Kicad render which doesn't show the soldermask opening for the P2 chip thermal bond to use the extra copper layers for additional heatsinking.
Those were worth showing. Us "veroboard" experimenters are reassured ;-) I guess there's one or two p1 products would benefit from a drop-in upgrade to p2 as well. The 0v bonds in the centre of the 100mil board look good for signal integrity.
Peter, which pins are available in the 40-pin DIP adapter, and what is the spacing between the 2 rows of pins? Will you sell a de-populated version of the P2D2 that just has the P2, crystal, voltage regulators and flash?
Peter, which pins are available in the 40-pin DIP adapter, and what is the spacing between the 2 rows of pins? Will you sell a de-populated version of the P2D2 that just has the P2, crystal, voltage regulators and flash?
For the original QUIP it's 1.2" on the inside row of 100mil pins since the rows of 50mil pins are 1" spaced. The DIP40 is of course 600mil as it should be
I have a few versions of the DIP40 adapters, since they are cheap to make. One has P0..P31 with the same pinout as the P1, while another has P0..P31 and P42..P47 since these pins are not committed on the P2PAL. The fact is, anyone can make a simple DIP adapter or anything else, and the P2D2 just plugs into it with cheap 50mil header pins or directly via SMD.
A lot of people think that these extras add a lot to the cost, but they don't. In fact it is cheaper to make these as one standard fully populated skinny module rather than having different versions. The only thing that may account for about $3 of the bom is the RTC itself, but it seems too good not to have it. As soon as I reduce the volume of any of these parts then they do become more costly and the base unit suffers from smaller volumes for each version and having to stock them. The clock gen adds $1, the USB adds about $1 including socket etc.
Bear in mind that while you can power this from the USB, it is not dependent upon USB and in fact the USB serial lines disconnect when the USB is not actively connected. This puppy is meant to be fully embedded but is also very happy as a tiny USB powered development system.
No special connectors or external USB is required to start playing with it.
For those who may wonder how the P2D2 core module mounts to the 100mil header plate, just think of it as another pcb layer where the two become one. Here is a quick 3d mesh of those 2 boards and the pin header spacing is exactly the same as it used to be. The 50mil holes are still there too except they moved in about 75mil to achieve an even 1" row to row spacing. This helps and allows room for the smd pads that mount the P2D2 to the "plate".
There's been a bit of talk about stripping away components to make P2 designs cheaper, but really the one big cost is the P2 itself. It is not a $2 or even $5 MCU, and it's a big chip with a big appetite. Now I could strip away the SD, the RTC, and USB from the P2D2, but which one would you buy, the stripped down one, or the one with the works for $5 extra? The only way to get the cost down is to remove the P2 chip or just standardize and put economy of volume to good work.
BOM costs of "extras" in 100 qty
EFM8UB30F40G-A-QFN20 (USB serial and reset control) = $1 (Mouser)
microUSB socket = 0.05 (LCSC)
microSD socket = =0.76 (Molex header) 0.10 (LCSC full socket)
RV-3028-C7 (RTC) = $2.05
Supercap = $1.18
TOTAL COST = $5.04 max
Sure, there's the 80 cent clock gen chip too that I could strip off too. It allows us to use 26MHz 0.5ppm GPS crystals and still get a standard 20MHz, or any other frequency that may be necessary that would be impractical otherwise.
The P2D2 is now a core module that can be configured as a big 100mil header board, or a DIP40, or a RPi HAT, or surface mounted directly to a board, or with cheap dual 50mil headers which I intend to use. Because it is a core module it doesn't have to change and so it can be manufactured in volume. As soon as I have different "models" then all the boards become more expensive, and which models do you stock etc?
There is a big difference in bom costs when dealing with 1k+ quantities too.
btw, the USB serial on-board DOES NOT compromise the design. You can power it externally and the USB will disconnect from the serial lines etc, but the power/reset monitor/watchdog always works.
I have a target price in mind for the standard P2D2 core module, and I am hoping that will be as cheap as $35. If it isn't, then it will be in that range. That is also why those who prepaid will get the extras such as the P2PAL even though I don't have economy of volume yet and my costs are higher.
Save components is a wast of time! Every feature that exist can be used. An engineers hour is more the 50€. And a missing component can lead to anger, violent, rioting, looting etc. So please calm down, use the chances P2D2 offers for free, distance socially and mask a bit. Send a free P2D2 to we all the people and they will stay at home and there will be no need for complimentary antibodies.
Narrow spacing is a plus but not with narrow minding.
If we don't save the world, who else ;-)
You could in future always offer a budget version with parts left depopulated to reduce costs if there are major pressures there but whatever you do don't attempt to redesign/reduce it now when you seem almost ready to ship something. Get the first versions out and test your market. For anyone who wants to buy volume quantities down the track but the price is unappealing you could offer a budget version without RTC/SD socket etc to shave a few $ if people don't always need it and that becomes a sticking point, but as you say the lion's share is already the processor anyway.
I have these 5 cent sd sockets that fit on the back, so that saves 70 cents. Woo Hoo, now you can go buy a ...... ummm..... I don't know what you can buy for 70 cents but I'm sure it's worth it!?
If I don't manufacture the ONE model then I won't get the cheaper price, and nobody will get it cheaper.
Please, do just one, fully featured model if that has the best performance to price ratio option. It is also the most desirable (by me at least) option. Then, we'll see what happens and then, having known the market response, there will be time for possible adjustments.
But please, start with what YOU think is the best option to start with.
It's all been set in FR4 as the best possible already.
All those extras for $5 is an amazing deal. I am sure that any applications that need the power of a P2 will most likely find a use for one or more of them. Impatiently waiting for mine to get here.
Here's a $10 K210 based CPU+WiFi module (placed over a board for comparison) that has dual core 64-bit 400MHZ RISC-V + AI +wifi+8MB ram etc. I might use it on another version of the p2pal to do both hd video and wifi and more. Probably runs micropython really well too.
Yeah, some more teasing .
Two more posts and we're going to see you have written a forth for this thing already (funny, I have it in my Mouser cart for a week now, just haven't pulled the trigger yet).
Comments
Is there a way to one's own post?
I'm finding out how much I can get the simple uPORT USB serial module assembled for and if the price is right I will sell them at the right price. Same goes for the P2D2 and the pricing will be aggressive, let's say, cost plus. The initial run will be done locally though.
There's the Moon festival in China at the moment adding a little delay to getting stuff back but in the meantime I have imported my files into Kicad and have even played with getting the 3D models working (mostly). Here is the P2D2 mostly right (the outside row of holes are castellated). (P2D2 QUICK REFERENCE)
Please do not forget us, poor Europeans, and get in touch with @ErNa too
PROTOTYPES
If all goes well my P2D2 boards and sundries will be here in a few days and I will be busy building up a tester. If I'm happy with the design and how the boards go together I will make up all I can with the stock I have but rather than ordering more parts at the moment I will get quotes on A&T from JLCPCB as they can also supply most of the parts from their LCSC division. Maybe I should give a US source a chance here too.
DESIGN
As you have seen, my designs are based on practical and commercial use, not educational nor hobbyist. Neither are they necessarily plug-in breadboard friendly although one can created a special breadboard JM style. My software is not purely for software purposes, but to run the hardware. My hardware is not ignorant of the software that runs on it either and even the layout reflects how I envisage the software will use it. P2D2 is a holistic design and while it may also look "pretty", it is entirely practical and versatile. At 2.4sq" including connectors it is a very compact design especially considering it is not a "bare-bones" design, but is fully fleshed out, and even more so with the P2PAL "layer" added. Placed next to a 40-pin DIP P1, there is not a lot of difference.
MOUNTING
When it comes to mounting the P2D2 can be mounted smd style flat to the pcb or 50mil matrix board with a 4mm max height, but even with 50mil pin header socket on the P2D2 component sided and flip mounted into pin headers that height is only around 6mm or 1/4". Of course you can also use 100mil pin headers when the P2D2 is smd mounted to the header card that has the same pinout and spacing as the original P2D2r4 along with the RPi header.
USB & RESET
Some may think that having USB onboard is great and some may think it is a hindrance and can cause problems with resetting etc. While that may be true of typical Parallax designs, I would never have USB if I couldn't bypass it and if it reset inadvertently either. In fact the USB is only part of the function of the support micro that is powered directly from the P2 supply since reset control is of paramount importance in a commercial design that needs to run 24/7 in a hostile environment. Besides the guaranteed reset that comes from a passive pull-down, active pull-up on power-good approach, there is also the watchdog that can at least restart the P2 and also maintain a non-volatile watchdog timeout log for fault analysis. BTW, the P2D2 will not reset simply because of USB or terminal connection/disconnect, it will only reset with a valid serial loader pulse, and even then it makes sure the P2D2 is forced into serial load mode despite the boot Flash and SD. The USB and reset support firmware is upgradable over USB and configurable in-circuit too. There is never any need for manual dip switches or jumpers. You can also drive P63/62 directly and besides the current limit resistor that makes that possible, the UB3 also floats its transmit line if it is not actively connected to USB.
SUPPORT FEATURES
Don't forget that while the P2D2 is plus featured, that it is still an economic design. The tight and optimized 2-layer board is very cheap to make, the USB and reset support only adds around $1 to the cost, and the same too with the flexible Si5351 clocking features where we can use precision crystals such 26MHz GPS crystals and still have a standard 20MHz, or any odd frequency up to 200MHz in fact. Then there are the multiple SD card footprints for various positions on the board along with software control over SD power cycling, but not only that, it also has series termination resistors for optimized high-speed drive. Real products normally need some kind of RTC, that is why may MCUs these days have a dedicated RTC built on-board, but nowhere near as good as an external RTC. The RV-3028-C7 is a 40na standby champion, and extremely accurate.
POWER
As well as switching the 1.8V supply, the 3.3V supply is also switched but at a slightly higher voltage and fed through a dual LDO for clean 3.3V I/O power. Besides the efficiency benefits of this approach it also benefits external circuitry since the 3.6V is available and can be fed directly into external chips without problem, even LDO regulated if needed. In fact the HyperRAM and ESP32 are fed directly from the 3.6V on the P2PAL layer which is especially important when you consider that the ESP32 may need 500ma during transmit. If a P2 system needed 1A on the 3.3V supply and it only used linear regulation then that would amount to 1.7W of heat dissipation no matter what, even if series diodes etc were used. A 3.7V Lithium can be connected to the VIO and there is a back-feed diode that ensures that the switcher is powered so that it can supply the 1.8V.
OPTIONABLE & UPGRADABLE
There are so many options to the base P2D2 itself but with extra P2PAL layer it is effectively turned into PC module with HyperRAM and the ESP32 etc, yet this thin layer added directly to the back of the P2D2 adds only 4mm to the height with 3mm of that being the ESP32 itself, and 1mm for the layer. But other P2PAL versions could be developed that mesh with the P2D2 that might have more memory or even a dedicated high resolution video CPU that can directly address video memory thus offload the problem of hires/hicolor displays and provide true USB host ports.
DELIVERY
This is the part that may be more interesting than even the compelling features of the P2D2 itself. If all goes well then October will see all the preorders filled and this week too I will take the time to test and update the working P2D2 boards I have that I had earmarked as free evaluation boards to some but due to the earlier problems with finalizing the UB3 firmware, I've had to belay that action while I make sure that everything else is sorted out (please give me a gentle pm nudge to remind me please). During the next few weeks as I have previously mentioned I will also find out about getting these made in an SMT assembly house, perhaps even JLCPCB/LCSC themselves, in which case there will be plenty of stock which will be kept locally both here and in the USA and also Europe I would imagine. btw, beyond these early production boards I have a very low and attractive target price set in mind that will make it a no-brainer to use P2D2.
SUPPORT
Yes please. I need your support.
Thank you
Happy to help you try out the HyperRAM on your P2PAL, though now my driver is released it should be possible for others to mess about with it too.
The flipside SD header is kind of redundant now as I find that if I do place a socket on the other side, I'd rather have it enclosed which is possible with the 3rd footprint which accommodates a fully enclosed push-push style socket. There are also these cheap naked but enclosed uSD sockets that Cluso99 sent me some of and they happen to fit on the flipside header pads but as a fully protected socket. So I don't think I'd ever use an actual Molex uSD header in that position whereas on the component side it is a real space-saver.
SD cards and LEDs have various height choices.
You can also get right angle LEDs, and those can avoid the SD card covering the main LED light of a vertical LED.
I like these ones, bright, and a good price, and not too small... - you can do a PCB design that takes a selection (vert or right angle) , just get the right angle ones right
https://lcsc.com/search?q=LTST-S220
This is true too as I forget that these Molex headers are available in various heights for that reason. The right-angled LEDs can only be used when the footprint and the angle match the polarity. Mind you, the P2PAL layer could be used simply for the LED indicators and uSD pattern, minus the ESP32 and HyperRAM etc. That way there are 6 LEDs altogether plus the WS2812 as well.
I will try to do 3d view with it all mounted to a base pcb.
btw, I will probably ship the blank DIP-40 and the QUIP-80 (original 100mil style P2D2) pcbs with each P2D2 as standard accessories along with the header pins. Mount them whichever way you wish then. I had thought of doing a P2PAL type layer that had the DIP40 socket for a very low profile connection but that would have precluded the use of an actual P2PAL enhancement. With a separate adapter it becomes an easy matter to swap around the P2D2 etc.
btw, the original version was done in Protel but this is a Kicad render which doesn't show the soldermask opening for the P2 chip thermal bond to use the extra copper layers for additional heatsinking.
For the original QUIP it's 1.2" on the inside row of 100mil pins since the rows of 50mil pins are 1" spaced. The DIP40 is of course 600mil as it should be
I have a few versions of the DIP40 adapters, since they are cheap to make. One has P0..P31 with the same pinout as the P1, while another has P0..P31 and P42..P47 since these pins are not committed on the P2PAL. The fact is, anyone can make a simple DIP adapter or anything else, and the P2D2 just plugs into it with cheap 50mil header pins or directly via SMD.
A lot of people think that these extras add a lot to the cost, but they don't. In fact it is cheaper to make these as one standard fully populated skinny module rather than having different versions. The only thing that may account for about $3 of the bom is the RTC itself, but it seems too good not to have it. As soon as I reduce the volume of any of these parts then they do become more costly and the base unit suffers from smaller volumes for each version and having to stock them. The clock gen adds $1, the USB adds about $1 including socket etc.
Bear in mind that while you can power this from the USB, it is not dependent upon USB and in fact the USB serial lines disconnect when the USB is not actively connected. This puppy is meant to be fully embedded but is also very happy as a tiny USB powered development system.
No special connectors or external USB is required to start playing with it.
BOM costs of "extras" in 100 qty
EFM8UB30F40G-A-QFN20 (USB serial and reset control) = $1 (Mouser)
microUSB socket = 0.05 (LCSC)
microSD socket = =0.76 (Molex header) 0.10 (LCSC full socket)
RV-3028-C7 (RTC) = $2.05
Supercap = $1.18
TOTAL COST = $5.04 max
Sure, there's the 80 cent clock gen chip too that I could strip off too. It allows us to use 26MHz 0.5ppm GPS crystals and still get a standard 20MHz, or any other frequency that may be necessary that would be impractical otherwise.
The P2D2 is now a core module that can be configured as a big 100mil header board, or a DIP40, or a RPi HAT, or surface mounted directly to a board, or with cheap dual 50mil headers which I intend to use. Because it is a core module it doesn't have to change and so it can be manufactured in volume. As soon as I have different "models" then all the boards become more expensive, and which models do you stock etc?
There is a big difference in bom costs when dealing with 1k+ quantities too.
btw, the USB serial on-board DOES NOT compromise the design. You can power it externally and the USB will disconnect from the serial lines etc, but the power/reset monitor/watchdog always works.
I have a target price in mind for the standard P2D2 core module, and I am hoping that will be as cheap as $35. If it isn't, then it will be in that range. That is also why those who prepaid will get the extras such as the P2PAL even though I don't have economy of volume yet and my costs are higher.
Narrow spacing is a plus but not with narrow minding.
If we don't save the world, who else ;-)
If I don't manufacture the ONE model then I won't get the cheaper price, and nobody will get it cheaper.
But please, start with what YOU think is the best option to start with.
All those extras for $5 is an amazing deal. I am sure that any applications that need the power of a P2 will most likely find a use for one or more of them. Impatiently waiting for mine to get here.
And I hope we can move these things around Europe before UK leaves EU for good. Not much time left it seems.
Two more posts and we're going to see you have written a forth for this thing already (funny, I have it in my Mouser cart for a week now, just haven't pulled the trigger yet).